Apply Probabilizer function

This function applies a probabilizer model previously trained by a probabilizer.train procedure type, converting a score to a probability.

Input and Output Values

It has one real-valued input value named score, usually the output of a classifier function type. It outputs one value named prob which is a real value between 0 and 1 indicating the probability associated with the input score.

Configuration

A new function of type probabilizer named <id> can be created as follows:

mldb.put("/v1/functions/"+<id>, {
    "type": "probabilizer",
    "params": {
        "modelFileUrl": <Url>
    }
})

with the following key-value definitions for params:

See also

• The classifier.train procedure type trains a classifier.
• The classifier.test procedure type allows the accuracy of a predictor to be tested against held-out data.
• The probabilizer.train procedure type trains a probabilizer.
• The classifier function type applies a classifier to a feature vector, producing a classification score.
• The classifier.explain function type explains how a classifier produced its output.
• The probabilizer function type works with classifier.apply to convert scores to probabilities.